We describe the Behavioral Theory of timing (BeT); the theory holds that temporal judgments are conditional discriminations based on behaviora elicited by periodic reinforcements. Simple mathematical models are developed to map the distributions of those behaviors in time, and to make predictions about ensuing temporal judgments. The models are constrained by the empirical fact of scalar timing (i.e., Weber's law), as are models of an alternate theory, Scalar Expectancy Theory (SET). We posit a pacemaket-countersystem, but BeT proposes that the period of the pacemaker varies with the rate of reinforcement, whereas SET holds that it does not. BeT predicts a positive but imperfect correlation between multiple temporal judgments within an interbal; the basic version of SET predicts no correlation. BeT predicts a close correspondence between overt behaviors and temporal estimates. Set does not. BeT predicts that different drive states that support different constellations of asjunctive behaviors will be differentially effective in supporting temporal judgments. SET does not. Other aims of this proposal involve the further development of BeT. The greatest need is for better models for the pacemaker which permit imperfect periodicity, rather than perfect periodicity with noise (SET) or complete randomness (BeT's Poisson model) that are invoked by these theories. In particular, basic models of nonlinear oscillators will be developed, and experiments to constrain versions of these models will be generated. The proposed research will contribute to the mental health mission of NIH by providing a comprehensive understanding of the role of reinforcers and their distribution in time on behavior